DEV Community: An Pham

I built an open-source Trello where a custom Pi agent harness does the cards

An Pham — Sat, 10 Oct 2026 11:44:08 +0000

My to-do list kept growing faster than I could clear it, and most of it wasn't hard work. "Compare three suppliers." "Draft the follow-up email." "Check the site still loads." Any decent AI agent could do these. The annoying part was everything around them: open a chat, paste the context, wait, copy the answer somewhere, then forget what I'd asked two days later.

I wanted the cards to just get done. So I built OpenLeo, a to-do board where agents, running on my own harness built on Pi, pick up the cards. It's free and open source under AGPL-3.0.

Here's the 30-second version:

How it works

It looks and feels like Trello: lists, with cards in them. The one difference is that a list can have an agent.

On the board above, a card dropped into Research goes to the researcher agent and a card in Writing goes to the writer. The Checks list has a monitor that looks at my site every weekday at 9:00. Inbox has no agent, so those cards wait for me.

When an agent finishes, the answer stays on the card. Here's a real one. I asked for the price of a Hario V60 at three shops, and the researcher came back with a little table and links:

A card can start because I pressed run, because its schedule came round, or because it landed in a list with an agent. Each run gets noted on the card, so when I look the next morning I can see what happened and when.

The same board also works as a table, with whatever fields you add (I used priority, due date and budget):

Every board gets its own computer

I didn't want agents poking around my real laptop. In OpenLeo each board has its own Linux computer, a cua sandbox, and the agents do their work in there. They open a browser, click around, fill in forms and save files. If I'm curious I open the computer and watch, and if an agent gets stuck I can take over the mouse.

On a Mac it uses Apple's container runtime when it can and falls back to Docker otherwise, and on Windows and Linux it uses Docker. You can also run the computers in cua's cloud. On first launch OpenLeo walks you through setting one up.

The board's notes, its memory and the skills you teach an agent (how you file receipts, say) live on that computer too, right next to the work.

It runs on your ChatGPT plan

I really didn't want people fiddling with API keys. OpenLeo uses OpenAI's Sign in with ChatGPT for open-source apps, the same kind of login Codex uses. You sign in with your ChatGPT account and the agents use the plan you already pay for. The tokens sit in a local vault file with owner-only permissions and never get sent to the web page.

What it isn't

It's for one person and their agents. I'm deliberately not turning it into a team project-management tool. It's also early. Expect rough edges, and back up your data before updating.

Try it

The site is leooi.com and the code is at github.com/ancs21/openleo.

git clone https://github.com/ancs21/openleo
cd openleo
bun install
bun run dev   # http://127.0.0.1:3000

Or skip the terminal and get the desktop app. There are builds for macOS (Apple Silicon), Windows (x64) and Linux (x64 and arm64) on leooi.com/download and the releases page.

So here's my question for you: if a to-do card could do itself, which one would you hand off first? And which one would you never trust an agent with? I'll read every answer, and if something breaks, open an issue.

Your AI agent can't fetch behind logins. I built a <400kb fix in Zig.

An Pham — Sun, 29 Mar 2026 02:34:36 +0000

Claude Code has WebFetch. Every AI coding agent has some way to grab a webpage.

They all hit the same wall: auth.

Try fetching a Medium article. A Facebook post. A private Substack. Anything behind Cloudflare. The agent gets a login page, a CAPTCHA, or nothing — and makes up the rest.

What playpanda does

playpanda https://medium.com/@someone/private-post

You log in once. Cookies get saved. After that, fetching just works — from an agent, a script, or the terminal.

The output is markdown with the junk stripped out. No nav bars, no cookie banners, no ads, no image CDN URLs. Just the content.

How agents can use it

Anything that can shell out can use it:

# Replace WebFetch in Claude Code
playpanda https://docs.example.com/private-api

# Fetch multiple pages at once
playpanda https://url1.com,https://url2.com,https://url3.com

How fetching works

It tries the fastest method first and escalates when something gets in the way:

Method	Speed	When it kicks in
HTTP fetch	~200ms	Default, works for most pages
JS rendering	~1.5s	Page needs JavaScript to load
Real browser + anti-detection	~6s	Site actively blocks bots

The agent doesn't need to know which tier ran. It just gets markdown back.

Size

Single Zig binary, under 400kb.

Install

One-liner:

curl -fsSL https://raw.githubusercontent.com/ancs21/playpanda/main/scripts/install.sh | sh

As an agent skill:

npx skills add ancs21/playpanda

From source (requires Zig 0.15+):

git clone https://github.com/ancs21/playpanda.git
cd playpanda
zig build -Doptimize=.ReleaseFast
cp zig-out/bin/playpanda ~/.local/bin/

GitHub: https://github.com/ancs21/playpanda (Apache 2.0 license)

TIL: Croc - Securely transfer files via CLI

An Pham — Tue, 21 Dec 2021 15:09:31 +0000

Introduction

croc is a tool that allows any two computers to simply and securely transfer files and folders.

Install

macOS

brew install croc

Ubuntu

curl https://getcroc.schollz.com | bash

Window

scoop install croc

Usage

Send file or folder

croc send [file(s)-or-folder]

Receive the file

croc [code from output]

More feature, please visit to https://github.com/schollz/croc

TLDR: Database Naming

An Pham — Mon, 16 Aug 2021 15:16:27 +0000

Credit: Photo by Tobias Fischer on Unsplash

Table names should be plural (ex: users, customers,...)
Use underscore_names (ex: user_id, user_role,...)
Don't name primary key column id
Don't make column names unclear (ex: u_role,...)
Name foreign key columns the same as the columns they refer to